Journal of Marketing Research publishes the definitive GLP-1 grocery study

Researchers at Cornell University published a study on Dec. 18 in the Journal of Marketing Research linking survey answers about GLP-1 use to actual grocery and restaurant receipts from tens of thousands of US households. Within six months of starting a GLP-1 medication, households cut grocery spending by an average of 5.3% ^[1][3].

The paper, titled "The No-Hunger Games: How GLP-1 Medication Adoption is Changing Consumer Food Demand," was written by assistant professor Sylvia Hristakeva and professor Jura Liaukonyte of the Charles H. Dyson School of Applied Economics and Management ^[1][3]. Rather than asking people what they ate, the authors used transaction data from Numerator, a market research firm that tracks grocery and restaurant purchases for a nationally representative panel of about 150,000 households, then matched those records to repeated surveys asking whether anyone in the household was taking a GLP-1, when they started and why ^[1][2]. That let the team compare adopters with similar households that did not use the drugs ^[1][2].

What the spending data showed

The average 5.3% grocery decline was not spread evenly. Higher-income households cut grocery spending by more than 8% ^[1][2][3]. Spending at limited-service restaurants — fast-food chains, coffee shops and similar places — fell by about 8% ^[1][3].

Inside the cart, calorie-dense and ultra-processed items dropped most. Savory snacks fell roughly 10%, with similarly large declines in sweets, baked goods and cookies ^[1][2]. Even staples such as bread, meat and eggs declined ^[1][2]. Only a few categories rose: yogurt increased the most, followed by fresh fruit, nutrition bars and meat snacks ^[1][2]. "The main pattern is a reduction in overall food purchases," Hristakeva said. "Only a small number of categories show increases, and those increases are modest relative to the overall decline" ^[1][3].

Among households that kept using the medication, lower food spending lasted at least a year, though the size of the reduction shrank over time ^[1][2][3]. Roughly one-third of users stopped taking the drug during the study period, and their food spending returned to pre-adoption levels ^[1][2]. Cornell reported that after stopping, grocery baskets became slightly less healthy than before people started, partly because of increased spending on categories such as candy and chocolate ^[1]. "After discontinuation, the effects become smaller and harder to distinguish from pre-adoption spending patterns," Hristakeva said ^[1][2][3].

The study also tracked how common these drugs have become. The share of US households reporting at least one GLP-1 user rose from about 11% in late 2023 to more than 16% by mid-2024 ^[1][2]. Weight-loss users skewed younger and wealthier, while people taking the drugs for diabetes were older and spread more evenly across income groups ^[1][2]. A separate estimate from market researcher Circana puts the current figure at 23% of US households having someone using a GLP-1, and projects that by 2030 GLP-1 households will account for 35% of all food and beverage units sold ^[2]. The two figures cover different time periods and come from different panels, so they are not directly comparable.

Why it matters for patients

For people taking or considering a GLP-1, the study puts numbers on something many users describe anecdotally: appetite changes show up in the household budget, not just on the scale. A 5.3% cut in grocery spending is an average across many households, so individual experience will vary widely — and the authors compared groups, not individuals ^[1][2].

The pattern after stopping is notable. The reversion in spending, including a shift back toward candy and chocolate, suggests the purchasing changes are tied to being on the medication rather than to permanent new habits ^[1]. The authors caution that the study cannot fully separate the biological effects of the drugs from other lifestyle changes people make at the same time, though evidence from clinical trials plus the reversion after discontinuation points to appetite suppression as a likely mechanism ^[1].

This is spending data, not nutrition or health-outcome data. The study does not measure calories consumed, weight change, nutrient intake or medical results, and it does not identify which specific products — semaglutide (Ozempic, Wegovy, Rybelsus), tirzepatide (Mounjaro, Zepbound) or others — drove the effects. Those details are not in the published summaries.

What happens next

The paper appeared online Dec. 18, 2025; Cornell issued its news release on Dec. 22 ^[1][3]. The researchers say widespread adoption could push food manufacturers, restaurants and retailers to change package sizes, product formulations and marketing ^[1][2]. "At current adoption rates, even relatively modest changes at the household level can have meaningful aggregate effects," Hristakeva said ^[1]. Whether shelves and menus actually shift, and how fast, is not yet known.